A large-slant-look TOPS model ground plane BP self-focusing method based on improved spectral compression

Abstract

The application discloses a large-oblique TOPS mode ground plane BP self-focusing method based on improved spectrum compression, and comprises the following steps: representing a BP image in a range frequency domain, obtaining a spectrum compression function by using a stationary point principle; obtaining a spectrum support region center of a certain point in a scene based on the spectrum compression function, and replacing radar positions of each time of the spectrum support region center of the certain point in the scene with a beam rotation center to obtain a new spectrum support region center; obtaining a final spectrum compression function of a TOPS mode based on a beam rotation center of a frequency domain mode introduced by the new spectrum support region center; converting the BP image to an azimuth frequency domain based on the final spectrum compression function of the TOPS mode to obtain an error function, and obtaining a SAR image of the TOPS mode finally completing self-focusing and sidelobe suppression based on the error function and an azimuth window function. The application greatly reduces the operation complexity of a BP algorithm processing TOPS mode data.

Description

Technical Field

[0001] This invention belongs to the field of radar signal processing technology, specifically relating to a ground-plane BP self-focusing method for large-slant-look TOPS mode based on improved spectral compression. Background Technology

[0002] Large squint (slant-out) Terrain Observation by Progressive Scans (TOPS) mode synthetic aperture radar (SAR) imaging overcomes beamwidth limitations by jointly adjusting the beam attitude in the azimuth and range directions, acquiring scene information far exceeding the instantaneous illumination range in a short time. While achieving wide-swath imaging, TOPS mode ensures the imaging results achieve a predetermined azimuth resolution through beam scanning speed control, and its data acquisition method is more flexible compared to other imaging modes. Traditional range-Doppler and other frequency domain algorithms and their improvements, in the case of large squint, suffer from severe two-dimensional spatial variation in the image due to the numerous approximations used in their slant-range model and filtering process. This makes it difficult to obtain high-quality SAR images when processing TOPS mode data with severe azimuth spectral aliasing. Furthermore, frequency domain algorithms are limited by their two-dimensional imaging plane, only able to operate parallel to the data acquisition plane; additional interpolation operations are required to obtain the ground projection image. Time-domain algorithms, such as back projection (BP) imaging, can address the aforementioned issues when the system's pulse repetition frequency (PRF) requirement is not high. The BP algorithm yields SAR images with optimal focusing, but at the cost of high computational complexity. The Fast Factorized Back Projection (FFBPA) algorithm, through multi-stage fusion, sacrifices some imaging performance but reduces the computational complexity of time-domain algorithms. However, due to its extensive use of Fast Fourier Transform (FFT) operations, it is difficult to adapt to platforms such as Field Programmable Gate Arrays (FPGAs). Therefore, the appropriate time-domain algorithm should be selected flexibly based on the specific hardware platform used.

[0003] Furthermore, how to achieve sidelobe suppression and resolve the residual phase error after inertial navigation / GPS compensation are key issues in the temporal domain imaging of large-slant-out TOPS mode. Currently, there is no full-aperture processing method for sidelobe suppression in TOPS mode; for phase error estimation and compensation, the mainstream approach is autofocus algorithms represented by Phase Gradient Autofocus (PGA). Existing combinations of TOPS mode and autofocus algorithms are based on sub-aperture operations, requiring error estimation and fitting compensation after sub-aperture imaging, followed by re-imaging. The computational performance of existing hardware platforms cannot achieve real-time imaging using two temporal domain algorithms. Additionally, due to beam scanning in TOPS mode, the original resolution two-dimensional imaging grid cannot satisfy the sampling law, resulting in severe azimuth spectral aliasing in full-aperture BP images. Therefore, if a spectral compression-based BP image autofocus algorithm is to be used in TOPS mode, the azimuth blocks need to be approximated as a beam-focusing mode for autofocus processing and stitching. However, this approach leads to stitching seams in the compensated image. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, this invention provides a ground-plane BP self-focusing method based on improved spectral compression for large-angle TOPS modes. The technical problem to be solved by this invention is achieved through the following technical solution:

[0005] A large-slant-angle TOPS mode land-plane BP autofocusing method based on improved spectral compression, the large-slant-angle TOPS mode land-plane BP autofocusing method comprising:

[0006] S1. Obtain the large squint TOPS mode imaging model in the rectangular coordinate system;

[0007] S2. Based on the large squint TOPS mode imaging model, the radar transmits a signal, receives and demodulates it, and pulses it to obtain the echo signal. The echo signal is then projected back onto each pixel in the imaging area, and the imaging area is divided into a grid. For each grid point in the grid, the complex value of the grid point in the image is obtained through azimuth coherent accumulation to obtain the BP image.

[0008] S3. Represent the BP image in the distance-frequency domain and obtain the spectral compression function using the principle of stationary phase points;

[0009] S4. Based on the spectral compression function, obtain the center of the spectral support region at a certain point in the scene, and replace the radar position of the center of the spectral support region at each time with the beam rotation center to obtain a new center of the spectral support region.

[0010] S5. Based on the new spectral support region center, the beam rotation center of the frequency domain mode is introduced to obtain the final spectral compression function of the TOPS mode.

[0011] S6. Based on the final spectral compression function of TOPS mode, the BP image is converted to the azimuth frequency domain to obtain the error function, and based on the error function and the azimuth windowing function, the final TOPS mode SAR image with autofocus and sidelobe suppression is obtained.

[0012] In one embodiment of the present invention, the echo signal is:

[0013]

[0014] Where S represents the echo signal, v represents the radar velocity, and t represents the radar speed. a Indicates azimuth time, L represents synthetic aperture length, B represents signal bandwidth, and t r R(t) represents the time interval between time intervals. a ) represents t a The slant distance at time t, where c represents the speed of light, λ represents the wavelength, and R(t) a x0, y0) represents the slant distance of grid point Q(x0, y0, 0) in the large strabismus TOPS mode imaging model. In one embodiment of the present invention, the BP image is:

[0015]

[0016] Where I(x,y) represents the BP image, t s Indicates the start time of the aperture, t e S(R(t) represents the final time of aperture termination. a ,x0,y0),t a ) indicates the echo signal.

[0017] In one embodiment of the present invention, step S3 includes:

[0018] S3.1 Represent the BP image in the range-frequency domain, that is:

[0019]

[0020] Where I(x,y) represents the BP image, t s Indicates the start time of the aperture, t e K represents the final time of aperture termination. r K is the distance wavenumber. rmin K represents the minimum distance wavenumber. rmax S(R(t) represents the maximum value of the distance wavenumber. a ,x0,y0),t a R(t) represents the echo signal. a (x, y) represents the slant distance of the grid point (x, y) in the large strabismus TOPS mode imaging model, and t a Indicates location and time.

[0021] S3.2. Solve the BP image in the range-frequency domain using the stationary phase point principle to obtain the wavenumber domain vector. The wavenumber domain vector is:

[0022]

[0023] Among them, (K) x K y R(t) represents the wavenumber coordinate. a x0, y0) represents the slant range of grid point Q(x0, y0, 0) in the large squint TOPS mode imaging model, and the radar location is represented as (X(t)). a ),Y(t a ),Z(t a ));

[0024] S3.3. The spectral compression function is obtained from the vector and wavenumber center in the wavenumber domain. The spectral compression function is:

[0025]

[0026] Where, f(x) i ,y i ) represents the spectral compression function, x i and y i K represents the grid sampling points of the BP image in the azimuth and range directions, respectively. rc Indicates the wavenumber center.

[0027] In one embodiment of the present invention, step S4 includes:

[0028] S4.1. Based on the spectral compression function, the center of the spectral support region at a certain point in the scene is obtained through wavenumber decomposition. The center of the spectral support region at a certain point in the scene is:

[0029]

[0030] Among them, K rc Denotes the wavenumber center, K rc K is obtained by decomposing along the x and y directions on the ground plane. xc and K yc Two direction vectors are used to obtain the center of the spectral support region (K). xc ,K yc );

[0031] S4.2. Combining the wavenumber characteristics of the TOPS mode, the radar position at each time step of the spectral support region center of a certain point in the scene is replaced with the beam rotation center to obtain a new spectral support region center, which is represented as:

[0032]

[0033] Among them, (X) rot ,Y rot Z rot () represents the beam rotation center, and the new spectral support region center is (K). xi ,K yi ).

[0034] In one embodiment of the present invention, step S5 includes:

[0035] S5.1, The objective function f(x,y) at grid point (x i ,y i The two-dimensional partial derivatives at point () are:

[0036]

[0037] S5.2. Integrate the distance and orientation dimensions obtained in step S5.1 to obtain the spectral compression function in TOPS mode. The spectral compression function in TOPS mode is:

[0038]

[0039] S5.3 Correcting the spatially varying spectral tilt of the spectral compression function in TOPS mode to obtain the final spectral compression function of TOPS mode. The spectral tilt correction part is as follows:

[0040]

[0041] Wherein, g(x) i ,K y ) indicates the corrected spectral tilt.

[0042] In one embodiment of the present invention, step S6 includes:

[0043] S6.1 The final spectral compression function based on TOPS mode transforms the BP image into the azimuth frequency domain using azimuth FFT to obtain the motion error, and then derives a compensation function based on the motion error. The compensation function is as follows:

[0044]

[0045] Wherein, I′(K) x ,y i ) represents the motion error, (K) x K y (X) represents the wavenumber coordinates. rot ,Y rot Z rot ) indicates the beam rotation center. t represents the estimated motion error. a Indicates location and time.

[0046] S6.2. By performing an azimuth-to-IFFT transformation, the error function and the azimuth-to-windowing function are transformed to the azimuth time domain, resulting in the final TOPS mode SAR image with autofocus and sidelobe suppression. The TOPS mode SAR image is as follows:

[0047] I final (x,y)=ifft(I′(K x ,y i win a (K x ))

[0048] Among them, I final (x,y) represents the final TOPS mode SAR image after autofocus and sidelobe suppression. The TOPS mode SAR image, win a (K x ) represents the azimuth window function, and ifft represents the inverse fast Fourier transform.

[0049] The beneficial effects of this invention are:

[0050] (1) The advantage of the BP algorithm is that it directly obtains the image through distance-related coherent accumulation, and it is applicable to different imaging modes. For the TOPS mode, since the beam scans in the azimuth direction, only a portion of the grid is illuminated at the same azimuth time. This invention utilizes this characteristic to adaptively select the effective projection grid at the current time, effectively reducing the overall computational load of the algorithm.

[0051] (2) The present invention arranges the grid as follows: the grid center is set as the projection of the radar beam center on the ground at the azimuth center time, the beam direction on the ground at the azimuth center time is the range direction of the grid, and the direction perpendicular to the beam direction in the imaging plane is the azimuth direction. This arrangement enables the azimuth spectrum of the BP image to be parallel to the azimuth direction, and the subsequent spectrum compression process can be simplified under this arrangement.

[0052] (3) This invention improves the spectral compression function by introducing the concept of beam rotation center in the frequency domain mode and extends it to the TOPS mode. It can directly compress the spectrum of the TOPS mode BP image with full aperture, making full aperture autofocus and sidelobe suppression possible, and greatly improving the imaging performance of BP real-time processing of TOPS mode. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating a ground-plane BP self-focusing method based on improved spectral compression in a large-angle TOPS mode according to an embodiment of the present invention.

[0054] Figure 2This is a flowchart illustrating another large-angle TOPS mode ground plane BP self-focusing method based on improved spectral compression provided in an embodiment of the present invention;

[0055] Figure 3 This is a schematic diagram of a large strabismus TOPS mode imaging model provided in an embodiment of the present invention;

[0056] Figure 4 This is a schematic diagram of a BP imaging grid arrangement provided in an embodiment of the present invention;

[0057] Figure 5 This is a schematic diagram illustrating the change of the wavenumber support region during spectral compression, provided by an embodiment of the present invention.

[0058] Figure 6 This is a schematic diagram of a random three-dimensional position error added in a simulation according to an embodiment of the present invention;

[0059] Figure 7 This is a comparative schematic diagram of the two-dimensional spectrum of a BP image under different conditions provided by an embodiment of the present invention;

[0060] Figure 8 This is a schematic diagram of BP imaging results based on simulation data of a large strabismus TOPS mode provided in an embodiment of the present invention;

[0061] Figure 9 This is a contrast diagram and azimuth sidelobe comparison diagram of point target 1 and point target 2 before and after motion compensation provided in an embodiment of the present invention;

[0062] Figure 10 This is a schematic diagram of the result after adding a two-dimensional window to a point target 1, as provided in an embodiment of the present invention. Detailed Implementation

[0063] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0064] Example 1

[0065] Please see Figure 1 and Figure 2 , Figure 1 This is a flowchart illustrating a large-slant-view TOPS mode ground-plane BP self-focusing method based on improved spectral compression, provided by an embodiment of the present invention. Figure 2 This is a flowchart illustrating another large-slant-view TOPS mode ground-plane BP autofocusing method based on improved spectral compression provided by an embodiment of the present invention. The embodiment of the present invention provides a large-slant-view TOPS mode ground-plane BP autofocusing method based on improved spectral compression, which includes steps S1-S6, wherein:

[0066] S1. Obtain the large squint TOPS mode imaging model in the rectangular coordinate system.

[0067] In this embodiment, please refer to Figure 3 The TOPS mode imaging model for large strabismus is as follows:

[0068] Taking the projection of the radar's azimuth center position onto the ground as the origin of the coordinate system, let's assume the radar beam illuminates ground target point p at the azimuth center time. a The azimuth time is given by the radar altitude H, and its three-dimensional coordinates are represented as (X(t)). a ),Y(t a For any point Q(x0,y0,0) in the scene, its slant distance can be expressed as:

[0069]

[0070] Wherein, R(t) a ;x0,y0) represents the slant distance of grid point Q(x0,y0,0) in the large strabismus TOPS mode imaging model.

[0071] S2. Based on the large squint TOPS mode imaging model, the radar transmits a signal, receives and demodulates it, and pulses it to obtain the echo signal. The echo signal is then projected back onto each pixel in the imaging area, and the imaging area is divided into a grid. For each grid point, the complex value of the grid point in the image is obtained through azimuth coherent accumulation to obtain the BP image.

[0072] In one specific embodiment, step S2 may include:

[0073] The radar transmitted signal, after reception, demodulation, and pulse compression, is represented as follows:

[0074]

[0075] Where S represents the echo signal, v represents the radar velocity, and t represents the radar speed. a Indicates azimuth time, L represents synthetic aperture length, B represents signal bandwidth, and t r R(t) represents the time interval between time intervals. a ) represents t a The slant distance at time t, where c represents the speed of light, λ represents the wavelength, and R(t) a ;x0,y0) represents the slant distance of grid point Q(x0,y0,0) in the large strabismus TOPS mode imaging model.

[0076] For point Q, its complex value in the image can be obtained through azimuth coherent accumulation. By performing the above operation on each grid point, the BP image can be obtained using the BP algorithm.

[0077]

[0078] Where I(x,y) represents the BP image, t s Indicates the start time of the aperture, t e S(R(t) represents the final time of aperture termination. a ,x0,y0),t a ) indicates the echo signal.

[0079] In this embodiment, the grid is arranged as follows: the grid center is set as the projection of the radar beam center onto the ground at the azimuth center moment; the beam direction at the azimuth center moment, along the ground projection direction, is the range direction of the grid; and the direction perpendicular to the beam direction in the imaging plane is the azimuth direction. Please refer to [link to relevant documentation]. Figure 4 , Figure 4 This is a schematic diagram of a BP imaging grid arrangement provided by an embodiment of the present invention. Such an arrangement enables the azimuth spectrum of the BP image to be parallel to the azimuth direction, and the subsequent spectral compression process can be simplified under this arrangement.

[0080] S3. Represent the BP image in the distance-frequency domain and obtain the spectral compression function using the principle of stationary phase points.

[0081] In one specific embodiment, step 3 may include:

[0082] S3.1 The purpose of spectral compression is to reduce the spectral range so that the image only needs a lower two-dimensional resolution to satisfy the sampling law. However, the original spectral compression is only applicable to spotting mode. Therefore, the BP image is represented in the range frequency domain, i.e.:

[0083]

[0084] Where I(x,y) represents the BP image, t s Indicates the start time of the aperture, t e K represents the final time of aperture termination. r K is the distance wavenumber. rmin K represents the minimum distance wavenumber. rmax S(R(t) represents the maximum value of the distance wavenumber. a ,x0,y0),t a R(t) represents the echo signal. a (x, y) represents the slant distance of the grid point (x, y) in the large strabismus TOPS mode imaging model, and t a Indicates location and time.

[0085] S3.2. Solve the BP image in the range-frequency domain using the stationary phase point principle (POSP) to obtain the wavenumber domain vector. The wavenumber domain vector is:

[0086]

[0087] Among them, (K) x K y () represents the wavenumber coordinates, and the radar location is represented as (X(t)). a ),Y(t a ),Z(t a )).

[0088] S3.3. The spectral compression function is obtained from the vector and wavenumber center in the wavenumber domain. The spectral compression function is:

[0089]

[0090] Where, f(x) i ,y i ) represents the spectral compression function, x i and y i K represents the grid sampling points of the BP image in the azimuth and range directions, respectively. rc Indicates the wavenumber center.

[0091] Please see Figure 5 , Figure 5 This is a schematic diagram illustrating the change of the wavenumber support region during spectral compression, provided by an embodiment of the present invention. Figure 5 (a) represents the original wavenumber support region. Figure 5 (b) is the wavenumber support region after the first step of spectral compression. Figure 5 (c) represents the wavenumber support region after the second step of spectral compression.

[0092] In this embodiment, traditional spectral compression functions cannot directly process TOPS mode data in real time. Previously, the sub-aperture method for estimating errors was relatively inefficient due to the short azimuth data. However, the sub-aperture method required two imaging operations: one for error estimation and another after stitching errors were compensated, significantly increasing time consumption. Furthermore, the method of directly dividing the BP image into azimuth blocks and constructing separate spectral compression functions, thus approximating each block as a clustered mode, suffers from discontinuous spectral compression function values ​​between blocks. This results in suboptimal compression in some areas and discontinuous energy at points along block boundaries, greatly reducing imaging performance.

[0093] S4. Based on the spectral compression function, obtain the center of the spectral support region at a certain point in the scene, and replace the radar position of the center of the spectral support region at each time with the beam rotation center to obtain a new center of the spectral support region.

[0094] Specifically, wavenumber decomposition can be used to obtain the center of the spectral support region at a certain point in the scene. By combining the wavenumber characteristics of the TOPS mode, the radar position of the center of the spectral support region at each time point in the scene can be replaced with the beam rotation center, resulting in a new expression for the center of the spectral support region.

[0095] In this embodiment, to ensure the real-time performance of the algorithm, the entire processing flow can only accommodate the computational load of a single BP imaging operation with a lower-resolution grid arrangement. Error estimation and compensation are crucial steps, but the autofocus algorithm cannot be applied to cases of spectral aliasing. Therefore, to solve this fundamental problem, a method capable of compressing the TOPS mode spectrum across the entire aperture is needed.

[0096] In one specific embodiment, step 4 may include:

[0097] S4.1 Based on the spectral compression function, the center of the spectral support region at a certain point in the scene is obtained through wavenumber decomposition.

[0098] Specifically, the actual beam change during data acquisition is a continuous process, occurring at different azimuth times t. a The center of the spectral support region at a point in the scene can be obtained through wavenumber decomposition, as shown below:

[0099]

[0100] Among them, K rc Denotes the wavenumber center, K rc K is obtained by decomposing along the x and y directions on the ground plane. xc and K yc Two direction vectors are used to obtain the center of the spectral support region (K). xc ,K yc ).

[0101] k rc It is caused by carrier frequency f c For the sub-aperture algorithm, the spectral compression function uses the radar position at the aperture center time as an approximation to replace the other positions, regardless of whether beam scanning exists. Because the sub-aperture time is extremely short, at sub-aperture T... a The radar position and beam direction change very little within a short period of time, so the error caused by this approximation can be ignored.

[0102] S4.2. Combining the wavenumber characteristics of the TOPS mode, the radar position of the center of the spectral support region at each time point in the scene is replaced with the beam rotation center to obtain a new center of the spectral support region.

[0103] Specifically, for the full-aperture focusing mode, the beam remains completely illuminating the target imaging area during long-term imaging, meaning that the x-axis involved in the imaging process within the synthetic aperture time is... py p The location will not change, but the radar position has changed significantly, so directly using the spectral compression function would be problematic.

[0104] Suppose that for each azimuth time, the phase error of directly using f(x,y) is... in It is different for each grid location. Because It happens to be about the azimuth center time t c Conjugate symmetric, phase error after coherent accumulation Conjugate cancellation yields a constant phase term, which has no effect on spectral compression and imaging results and can be ignored. That is, the spectral compression function obtained through azimuth coherent accumulation in the full-aperture focused imaging mode is exactly the same as the spectral compression function in step 3.3. Therefore, the full-aperture focused imaging mode can successfully complete the spectral compression operation using the spectral compression function.

[0105] Based on the wavenumber characteristics of the TOPS mode, the spectral compression function is improved as follows:

[0106] During TOPS mode imaging, the radar and beam pointing are constantly moving, and the position of the beam-illuminated grid is constantly changing. However, if we connect the position of the beam center at each moment with the current radar position and find the intersection point, this point is the beam rotation center P. rot (X rot ,Y rot Z rot ).

[0107] Replace the radar position at each moment at the center of the spectral support region of a certain point in the scene with the beam rotation center P. rot This allows us to obtain a new way of representing the center of the spectral support region:

[0108]

[0109] Among them, (X) rot ,Y rot Z rot () represents the beam rotation center, and the new spectral support region center is (K). xi ,K yi ).

[0110] S5. Based on the new spectral support region center, the beam rotation center of the frequency domain mode is introduced to obtain the spectral compression function of the TOPS mode.

[0111] In this embodiment, by introducing the concept of beam rotation center in the frequency domain mode, the present invention improves the spectral compression function and extends it to the TOPS mode, which can directly compress the spectrum of the TOPS mode BP image across the entire aperture, making full aperture autofocus and sidelobe suppression possible, and greatly improving the imaging performance of real-time BP processing of the TOPS mode.

[0112] In one specific embodiment, step S5 may include:

[0113] S5.1, The objective function f(x,y) at grid point (x i ,y i The two-dimensional partial derivatives at point () are:

[0114]

[0115] S5.2. Integrate the distance and orientation dimensions obtained in step S5.1 to obtain the spectral compression function in TOPS mode. The spectral compression function in TOPS mode is:

[0116]

[0117] S5.3 Correct the spatially variable spectral tilt of the spectral compression function in TOPS mode to obtain the final spectral compression function of TOPS mode.

[0118] After obtaining the spectral compression function in TOPS mode through step S5.2, the spectral centerline of any target point in the scene can be expressed by the following formula:

[0119]

[0120] Among them, (K) x K y () represents wavenumber coordinates;

[0121] Construct a compensation function g(x) i ,K y The following partial derivative relationship exists.

[0122]

[0123] By integrating, the correction part g(x) for the spectral tilt can be obtained. i ,K y )for:

[0124]

[0125] Wherein, g(x) i ,K y ) indicates the corrected spectral tilt.

[0126] S6. Based on the final spectral compression function of TOPS mode, the BP image is converted to the azimuth frequency domain to obtain the error function, and based on the error function and the azimuth windowing function, the final TOPS mode SAR image with autofocus and sidelobe suppression is obtained.

[0127] Specifically, after the spectral compression step, the BP image is converted to the azimuth frequency domain by azimuth FFT, and the motion error can be estimated by autofocus. Range windowing is performed during pulse compression, and azimuth windowing is performed after motion error compensation. Finally, the TOPS mode SAR image with autofocus and sidelobe suppression is obtained by converting it to the azimuth time domain by azimuth IFFT.

[0128] In one specific embodiment, step 6 may include:

[0129] S6.1 The final spectral compression function based on TOPS mode converts the BP image to the azimuth frequency domain through azimuth FFT to obtain the motion error, and obtains the compensation function based on the motion error.

[0130] Specifically, after the spectral compression step, the azimuth spectrum of the TOPS mode is shifted to the aligned position, and the azimuth spectrum is no longer aliased. The BP image is then converted to the azimuth frequency domain using azimuth FFT (Fast Fourier Transform), allowing for motion error estimation via autofocus. Let the estimated motion error be... The compensation function can then be expressed as:

[0131]

[0132] Wherein, I′(K) x ,y i ) represents motion error. This represents the estimated motion error;

[0133] S6.2. The error function and the azimuth windowing function are transformed to the azimuth time domain by azimuth-to-IFFT transformation to obtain the final TOPS mode SAR image with autofocus and sidelobe suppression.

[0134] For TOPS mode, range windowing can be conveniently performed during pulse compression. Let the pulse-compressed signal be sig. RC The pulse pressure reference function is s ref The distance-to-window function is win. r (t r If the distance windowing is applied, then the distance direction can be represented as:

[0135] sig RC =ifft(fft(sig)*fft(s) ref )*fft(win r (tr )))

[0136] Among them, sig RC This indicates adding a window to the distance.

[0137] With the aid of spectral compression, the azimuth spectrum of the BP image is aligned, therefore azimuth windowing can be performed after motion error compensation. Let the azimuth windowing function be win. a (K x After azimuth windowing, the image is transformed to the azimuth time domain using azimuth IFFT to obtain the final TOPS mode SAR image with autofocus and sidelobe suppression. The TOPS mode SAR image is as follows:

[0138] I final (x,y)=ifft(I′(K x ,y i win a (K x ))

[0139] Among them, I final (x,y) represents the final TOPS mode SAR image after autofocus and sidelobe suppression. The TOPS mode SAR image, win a (K x ) represents the azimuth windowing function, and ifft represents the inverse fast Fourier transform.

[0140] Time-domain algorithms, such as BP, combined with TOPS imaging mode can obtain theoretically optimal quality wide-swath SAR images. However, how to perform motion error compensation and sidelobe suppression in real time after imaging remains an unsolved problem in practical applications. This method analyzes the essential process of BP image spectral compression and derives an accurate TOPS mode spectral compression function by introducing the concept of a virtual rotation center. This allows for full-aperture compression of the TOPS mode BP image spectrum after imaging. This method enables autofocus and azimuth sidelobe suppression even with sparsely arranged grids, requiring only a single imaging process, significantly reducing the computational complexity of the BP algorithm in processing TOPS mode data. Theoretical derivation and simulation data processing results verify the effectiveness of this imaging method.

[0141] This embodiment verifies the large-angle TOPS mode ground plane BP self-focusing method based on improved spectral compression provided in the above embodiment through simulation experiments.

[0142] I. Simulation Experiment

[0143] 1. Simulation conditions

[0144] This experiment uses simulated TOPS mode echo data with large squint to verify the effectiveness of the proposed algorithm by adding high-order random errors. The simulation parameters are shown in Table 1.

[0145] Table 1 SAR Simulation Parameters

[0146] Synthetic aperture time (s) 1 Bandwidth (MHz) 160 Carrier frequency (GHz) 16.1 Velocity (m / s) (70,0,0) PRF(Hz) 2000 Oblique angle (°) at azimuth center time 65 Pulse width (µs) 10 Height (km) 3000

[0147] To simulate motion errors during real platform flight, and incorporating actual inertial navigation parameters, this invention simulates a velocity error with zero mean and a standard deviation of 1.5 m / s, integrates the results to obtain a three-dimensional position error, and adds this error to the radar trajectory for backpropagation (BP) imaging. Please refer to [link to relevant documentation]. Figure 6 , Figure 6 This is a schematic diagram of a random three-dimensional position error added in a simulation provided by an embodiment of the present invention.

[0148] 2. Simulation Content and Result Analysis

[0149] Please see Figure 7 , Figure 7 This is a comparative schematic diagram of the two-dimensional spectrum of a BP image under different conditions provided by an embodiment of the present invention. Figure 7 (a) is the original two-dimensional spectrum. Figure 7 (b) is the two-dimensional spectrum using the spectral compression function. Figure 7 (c) is a two-dimensional spectrum using the spectral compression algorithm of this invention. Figure 7 (b) Although the spectrum was compressed to some extent, its azimuth direction was still aliased, and the autofocus algorithm could not be used. Figure 7 (c) The effectiveness of the spectrum compression algorithm of the present invention can be clearly seen. It can reduce the grid density while ensuring the azimuth resolution, thereby further reducing the amount of computation.

[0150] Please see Figure 8 , Figure 8 This is a schematic diagram of BP imaging results based on simulated data from a large strabismus TOPS mode, provided in an embodiment of the present invention. (The diagram is compared with...) Figure 8 (a) and Figure 8 (b) demonstrates that, based on the improved spectral compression method proposed in this invention, the self-focusing algorithm can be directly used for the azimuth spectrum of sparsely sampled BP images.

[0151] Please see Figure 9 , Figure 9 This invention provides a comparison of the contour map and azimuth sidelobe of point target 1 and point target 2 before and after motion compensation, according to an embodiment of the invention. Figure 9 Given Figure 8 The comparison between the top left and bottom right corners before and after self-focusing shows that these two points represent the points with the most severe spatial variation in error. Figure 9 (a)-9(d) and Figure 9(e)-9(h) show the contour maps and azimuth sidelobes of point targets 1 and 2 before and after autofocusing, respectively. Since the error has a negligible impact on the range focusing effect, these are omitted here. Figure 9 It can be seen that motion error has a significant impact on azimuth focusing performance, resulting in severe azimuth defocusing. The TOPS mode full aperture spectrum compression based on the algorithm of this invention, followed by autofocusing, effectively compensates for motion error and significantly improves azimuth focusing performance.

[0152] Please see Figure 10 , Figure 10 This is a schematic diagram of the result after two-dimensional windowing of point target 1 according to an embodiment of the present invention. Table 2 shows a comparison of the two-dimensional peak sidelobe ratio (PSLR) and integral sidelobe ratio (ISLR) before and after autofocusing and azimuth windowing of point target 1. The sidelobe suppression effect is almost consistent for each point in the scene. Figure 10 Table 2 demonstrates the effectiveness of the azimuth sidelobe suppression method based on the improved spectral compression algorithm of this invention.

[0153] Table 2. Self-focusing and front and rear sidelobe indices of point target 1 with azimuth windowing.

[0154]

[0155] This invention discloses a real-time BP imaging method for large-angle TOPS mode based on improved spectral compression. Large-angle TOPS mode SAR imaging is of great significance as it can flexibly acquire wide-field information in a short time. This invention introduces the concept of a virtual rotation center into the spectral analysis of BP images and proposes a real-time BP imaging algorithm for large-angle TOPS mode based on improved spectral compression. This algorithm can compress the spectrum of TOPS mode BP images across the entire aperture, enabling autofocus and azimuth sidelobe suppression even with sparsely arranged grids. Compared to traditional BP algorithms for processing TOPS mode data, this algorithm only requires one imaging process, significantly reducing the computational complexity of BP algorithms for processing TOPS mode data. Theoretical derivation and simulation data processing results verify the effectiveness of this imaging method.

[0156] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0157] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0158] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for ground plane BP autofocusing of large squint TOPS mode based on improved spectral compaction, characterized in that, The large-angle strabismus TOPS mode ground plane BP autofocusing method comprises: S1, acquiring a large-angle strabismus TOPS mode imaging model under a rectangular coordinate system; S2, based on the large-angle strabismus TOPS mode imaging model, receiving demodulation after radar signal transmission, and obtaining an echo signal through pulse compression, projecting the echo signal to each pixel in an imaging area in reverse, and dividing the imaging area into a grid, for each grid point in the grid, obtaining a complex value of the grid point in the image through azimuth coherence accumulation to obtain a BP image; S3, representing the BP image in a range frequency domain, and obtaining a spectrum compression function by using a stationary point principle; S4, obtaining a spectrum support area center of a certain point in a scene based on the spectrum compression function, and replacing radar positions of the spectrum support area center of the certain point in the scene at each time with a beam rotation center to obtain a new spectrum support area center; S5, obtaining a final spectrum compression function of the TOPS mode based on the beam rotation center of the frequency domain mode introduced by the new spectrum support area center; S6, converting the BP image to an azimuth frequency domain based on the final spectrum compression function of the TOPS mode to obtain an error function, and obtaining a TOPS mode SAR image in which autofocusing and sidelobe suppression are finally completed based on the error function and an azimuth windowing function.

2. The TOPS mode ground plane BP autofocusing method for large strabismus according to claim 1, characterized in that, The echo signal is: where S represents an echo signal, v represents a radar velocity, t a represents an azimuth slow time, L represents a synthetic aperture length, B represents a signal bandwidth, t r represents a range fast time, R(t a ) represents a slant range at t a , c represents a light speed, and λ represents a wavelength. R(t a ; x0, y0) represents a slant range of a grid point Q(x0, y0, 0) in a large squint TOPS mode imaging model.

3. The TOPS mode ground plane BP autofocusing method for large strabismus according to claim 2, characterized in that, The BP image is: where I(x, y) represents the BP image, t s represents the aperture start time, t e represents the aperture end time, S(R(t a , x0, y0), t a ) represents the echo signal.

4. The TOPS mode ground plane BP autofocusing method for large strabismus according to claim 1, wherein, Step S3 comprises: S3.1, representing the BP image in a range frequency domain, that is: where I(x, y) denotes the BP image, t s denotes the aperture start time, t e denotes the aperture end time, K r is the range wavenumber, K rmin denotes the minimum value of the range wavenumber, K rmax denotes the maximum value of the range wavenumber, S(R(t a ,x0,y0),t a ) denotes the echo signal, R(t a ,x,y) denotes the slant range of the grid point (x, y) in the wide- angle TOPS mode imaging model, t a denotes the azimuth slow time; S3.2, solving the BP image represented in the range frequency domain by using a stationary point principle to obtain a wave number domain vector, the wave number domain vector being: wherein (K x , K y ) represents the wave number coordinate, R(t a ; x0, y0) represents the slant range of the grid point Q(x0, y0, 0) in the large-oblique TOPS mode imaging model, and the position of the radar is represented as (X(t a ), Y(t a ), Z(t a )) S3.3, obtaining a spectrum compression function according to the wave number domain vector and a wave number center, the spectrum compression function being: where f(x i , y i ) denotes the spectral compaction function, x i and y i denote the grid sampling points of the BP image in azimuth and range respectively, and K rc denotes the wave number center.

5. The TOPS mode ground plane BP autofocusing method for large strabismus according to claim 4, characterized in that, Step S4 comprises: S4.1, obtaining a spectrum support area center of a certain point in a scene by wave number decomposition based on the spectrum compression function, the spectrum support area center of the certain point in the scene being: where K rc represents the wave number center, K rc represents the wave number center, K xc and K yc are two direction vectors, and K xc , K yc are the centers of the spectral support region. S4.2, replacing radar positions of the spectrum support area center of the certain point in the scene at each time with a beam rotation center in combination with a wave number characteristic of the TOPS mode to obtain a new spectrum support area center, the new spectrum support area center being represented as: where (X rot ,Y rot ,Z rot ) denotes the beam rotation center, and the new spectral support region center is (K xi ,K yi ).

6. The TOPS mode ground plane BP autofocusing method for large strabismus according to claim 5, wherein, Step S5 comprises: S5.1 The two-dimensional partial derivatives of the objective function f(x,y) at the grid point (x i ,y i ) are: S5.2, performing integral operation on the distance and azimuth two dimensions obtained in step S5.1 to obtain a spectrum compression function under the TOPS mode, the spectrum compression function under the TOPS mode being: S5.3, correcting a frequency spectrum tilt of the spectrum compression function under the TOPS mode, to obtain a final spectrum compression function of the TOPS mode, the frequency spectrum tilt correction part being: where g(x i ,K y ) represents the corrected spectral tilt.

7. The TOPS mode ground plane BP autofocusing method for large strabismus according to claim 1, wherein, Step S6 comprises: S6.1, converting the BP image to an azimuth frequency domain by azimuth FFT based on the final spectrum compression function of the TOPS mode to obtain a motion error, and obtaining a compensation function based on the motion error, the compensation function being: where I'(K x ,y i ) represents the motion error, (K x , K y ) represents the wave number coordinate, (X rot , Y rot , Z rot ) represents the beam rotation center, represents the estimated motion error, t a represents the azimuthal slow time; S6.2, converting the error function and the azimuth windowing function to an azimuth time domain by azimuth IFFT conversion to obtain a TOPS mode SAR image in which autofocusing and sidelobe suppression are finally completed, the TOPS mode SAR image being: I final (x,y) = ifft(I'(K x ,y i ) win a (K x )) where I final (x, y) represents the SAR image of the TOPS mode finally completed with autofocusing and sidelobe suppression, the SAR image of the TOPS mode, win a (K x ) represents the azimuth window function, and ifft represents the inverse fast Fourier transform.

Images